Marine and Underwater Science and Technology

Simulation and Dynamic Analysis

of Remotely Operated Vehicle (ROV)
Using PID Controller for Pitch Movement
†
Hujjatul Anam, ‡Joga Dharma Setiawan, and ‡Agus Budiyono
†
Bhimasena Research and Development, Sumedang, Indonesia
‡
Department of Mechanical Engineering, Diponegoro University, Semarang, Indonesia
‡
School of Aerospace, Mechanical and Manufacturing Engineering, Aerospace & Aviation Program, RMIT University, Australia

Abstract—ROV (Remotely Operated Vehicle) is a type of velocity plot with respect to time. The output of this research is
underwater robot that resembles a mini submarine. The robot is an ROV design that is based on mathematical model.
remotely controlled by a pilot. In this ROV tether is usually used to
link the robot with devices such as the sea surface energy source, A. ROV (Remotely Operated Vehicle)
remote control and sensing display. This study started from the stage A typical ROV consists of frame, thruster, control system/
of modeling the body of the ROV using SolidWorks. Estimation of
box, and camera.
added mass coefficient and hydrodynamics damping was performed
using first principle approach. This leads to development of the
equations of motion of ROV describing its salient dynamics. Based on
the mathematical model, the ROV motion in 3-D space was
investigated by using MATLAB SIMULINK. The stability of the
ROV was analyzed and a classical control design was performed using
the PID control toolbox.
Keywords—ROV, PID controller, virtual reality, dynamic model.

I. INTRODUCTION

A S we know, two-third of Indonesia’s area is covered by

sea. Being a maritime country, Indonesia’s sea area is no
less than 5.8 million km2, which consists of 0.8 million km2 of
territorial sea area, 2.3 million km2 of archipelagic sea, and 2.7
million km2 of exclusive economic zone [1]. Such a vast area is
awaiting to be explored and can be done economically and
safely by underwater robots. Based on operation system,
underwater robots are divided in two types: Remotely Operated
Vehicle (ROV), and Autonomous Underwater Vehicle (AUV).
ROV is underwater vehicle that is controlled by human directly
using remote control from surface, and is usually operated in
the deep sea and controlled from ship using tether. Whereas
AUV is underwater vehicle that can move without being
controlled by human operator; it can be operated autonomously
below the surface but not too deep [2].
In this research, simulation is made using toolbox sl3d in
MATLAB 2013b whereas the dynamic analysis of ROV uses
mathematical model and computational parameter. Simulation
using MATLAB Simulink is performed to obtain position and

Corresponding author: Agus Budiyono (e-mail: agus.budiyono@rmit.edu.au).

This paper was submitted on May 31, 2015; revised on April 3, 2015; and Figure 1 ROV model
accepted on May 20, 2015.
 Marine and Underwater Science and Technology 24

B. Coordinate System, Definition of Kinematics and Position II. DYNAMIC MODELING OF ROV
of ROV A. Kinematic Model of ROV
Coordinate system is important to explain the motion of Kinematics is part of dynamics, which is study about motion
underwater vehicle in six degrees of freedom (DOF) and to
of objects without regard to the forces that influence the motion
determine position and orientation in three dimensional volume
of the objects. Description of kinematics from ROV is based on
and time. The 6 DOF motion is described as position and
translational velocity where X, Y, Z, and ϕ, θ, ψ are for two types of reference, which is, body-fixed coordinate system
orientation and rotational velocity, respectively. For and earth-fixed coordinate system. Earth-fixed coordinate
underwater robots, those motion variables are termed as surge, system is used as the inertial reference with assumption that
sway, heave, roll, pitch, and yaw. Position or translational earth is superficial and its motion is neglected [4]. For object
motion and orientation or rotational motion of a rigid body (a moving in three dimensions, variables that must be considered
body in which the relative position of all its points is constant) in the kinematic model are:
can be described with respect to a reference position. For this
purpose, some set of orthogonal coordinate axes are chosen and η1 =[x y z]T (1)
assumed to be rigidly connected to the arbitrary origin of the η2 =[ϕ θ ψ]T (2)
body to build up the reference frame.
v1 =[u v w]T (3)
Force and moment that work in underwater vehicle must be
referenced to the same frame as the motion variables. The v2 =[p q r]T (4)
notation that is used to explain 6 degree of freedom (DOF) is
summarized in TABLE I. Note that convention of symbols for τ1 =[X Y Z]T (5)
underwater vehicle is: x as positive direction in the front, y as τ2 =[K M N]T (6)
positive direction in the right hand, z as positive direction in the
bottom and rule of right hand is used for angle [3]. vector 𝜂 is position and orientation coordinate about ROV in
TABLE I NOTATION OF STANDARD MOTIONS [3] the reference of earth-fixed coordinate system, whereas vector
𝑣 is velocity vector, and each 𝜏 respectively represents external
Linear and
Forces and Positions and force vector and external moment vector that acts on ROV
DOF Motions Angular
Moments Orientations
Velocities body. To obtain the kinematic expression of ROV, the
1 Surge X U x formulation must be transformed from body-fixed coordinate
system to earth-fixed coordinate system. The transformation
2 Sway Y V y
matrix is usually written as follows [4]:
3 Heave Z w z
4 Roll K p ϕ η̇ 1 =J1 (η2 )v1 (7)
5 Pitch M q θ ẋ u
6 Yaw N r ψ [ẏ ] =J1 (η2 ) [ v ] (8)
ż w
Figure 2 shows the determination of body-fixed coordinate
system and earth-fixed coordinate system that are used as
c ψcθ -s ψ c θ + c ψ s θ s ϕ s ψ s ϕ+ c ψ s θ c ϕ
standard motion as has been summarized in TABLE .
J1 (η2 )= [s ψ c θ cψcθ+sψsθsϕ -c ψ s ϕ+ s ψ s θ c ϕ] (9)
-sθ cθsϕ cθcϕ

-1 T
(J1 (η2 )) = (J1 (η2 )) (10)

ẋ =u c ψ c θ +v(-s ψ c ϕ + c ψ s θ s ϕ)+w(s ψ s ϕ+ c ψ s θ c ϕ) (11)

ẏ =u s ψ c θ +v(c ψ c ϕ + s ψ s θ s ϕ)+w(-c ψ s ϕ+ s ψ s θ c ϕ) (12)
ż =u (- s θ) +v(c θ s ϕ)+w( c θ c ϕ ) (13)

Transformation to earth-fixed coordinate system:

ϕ̇ p
[ θ̇ ] =J2 (η2 ) [q] (14)
ψ̇ r
Figure 2 Body-fixed and earth-fixed coordinate system
 Marine and Underwater Sci. and Tech., 2015, Vol. 1, No. 1

1 sϕtθ cϕtθ 𝐹𝑗 = −𝑈̇𝑖 𝑚𝑗𝑖 − 𝜀𝑗𝑘𝑙 𝑈𝑖 Ω𝑘 𝑚𝑙𝑖 (25)

J2 (η2 )= [0 cϕ -s ϕ ] (15)
0
sϕ cϕ 𝑀𝑗 = −𝑈̇𝑖 𝑚𝑗+3,𝑖 − 𝜀𝑗𝑘𝑙 𝑈𝑖 𝛺𝑘 𝑚𝑙+3𝑖 − 𝜀𝑗𝑘𝑙 𝑈𝑖 𝑈𝑘 𝑚𝑙𝑖 (26)
cθ cθ

ϕ̇ =p+q(s ϕ t θ)+r(c ϕ t θ) (16) where j, k, l, the value of 1, 2, and 3, while i is used as a six
speed components.
θ̇ =q c ϕ-r s ϕ (17)
Ui are the three components the speed of translational (surge,
ψ̇ =q s ϕ / c θ+r c ϕ/ c θ (18)
sway, heave) and Ω1=U4, Ω2=U5, Ω3=U6, are the three
components the speed of rotation (roll, pitch, yaw). Alternating
B. Dynamic Model of ROV Tensor +1 if the index is equal to the cyclic order (123, 231,
Based on Newton’s Second Law, derivation of the 312), equal to -1 if the index acyclic (132, 213, 321) and zero if
equations of motion will derive the complete the same index.
three-dimensional, six degree of freedom (DOF) rigid body Since the plane xz plane is symmetric (right/ leftside thruster
governing equations, that will briefly work through the symmetry), then the additional mass matrix is as follows:
derivation of the dynamic equation of the general form
𝐹 = 𝑚 ∙ 𝑎, where F is external force that is working on the 𝑋𝑢̇ 𝑋𝑣̇ 𝑋𝑤̇ 𝑋𝑝̇ 𝑋𝑞̇ 𝑋𝑟̇
central of mass of the objects (N), whereas 𝑚 ∙ 𝑎 is the product 𝑌𝑢̇ 𝑌𝑣̇ 𝑌𝑤̇ 𝑌𝑝̇ 𝑌𝑞̇ 𝑌𝑟̇
of mass (kg) and acceleration (m/s2). External forces consist of 𝑍𝑢̇ 𝑍𝑣̇ 𝑍𝑤̇ 𝑍𝑝̇ 𝑍𝑞̇ 𝑍𝑟̇
𝑀𝐴 = − (27)
radiation-induced, environmental and propulsive forces. 𝐾𝑢̇ 𝐾𝑣̇ 𝐾𝑤̇ 𝐾𝑝̇ 𝐾𝑞̇ 𝐾𝑟̇
Radiation-induced forces consist of hydrodynamics damping 𝑀𝑢̇ 𝑀𝑣̇ 𝑀𝑤̇ 𝑀𝑝̇ 𝑀𝑞̇ 𝑀𝑟̇
force and restoring force. Although derivation of dynamic [𝑁𝑢̇ 𝑁𝑣̇ 𝑁𝑤̇ 𝑁𝑝̇ 𝑁𝑞̇ 𝑁𝑟̇ ]
equation is taken from various references, in this paper,
summary of how to get the complete data for the motion Because the inspection position of the center point of the
dynamic equation of ROV is presented [5][6][7]. ROV, non-zero added mass can be reduced to a simpler matrix,
the matrix equation to obtain non-zero added mass for the
 Axial force equation (surge): ROV. ROV design assumptions of symmetry on the left and
∑ Xext =m [u̇ +wq-vr-xG (q2 + r2 )+ zG (pr+ q̇ )] (19) right side cross section as follows:

 Lateral force equation (sway): 𝑚11 0 𝑚13 0 𝑚15 0

0 𝑚22 0 𝑚24 0 𝑚26
∑ Yext =m [v̇ -wp+ur+xG (pq+ ṙ )+ zG (qr- ṗ )] (20) 𝑚 0 𝑚33 0 𝑚35 0
𝑀𝐴 = − 31 (28)
0 𝑚42 0 𝑚44 0 𝑚46
𝑚51 0 𝑚53 0 𝑚55 0
 Normal Force equation (heave):
[0 𝑚62 0 𝑚64 0 𝑚66 ]
𝑋𝑢̇ 0 𝑋𝑤̇ 0 𝑋𝑞̇ 0
∑ Zext =m [ẇ -uq+vp+xG (pr- q̇ ) - zG (q2 + p2 )] (21)
0 𝑌𝑣̇ 0 𝑌𝑝̇ 0 𝑌𝑟̇
𝑍𝑢̇ 0 𝑍𝑤̇ 0 𝑍𝑞̇ 0
 Rolling moment equation: 𝑀𝐴 = − (29)
0 𝐾𝑣̇ 0 𝐾𝑝̇ 0 𝐾𝑟̇
∑ Kext =Ixx ṗ -Ixy +(Izz -Iyy )qr-m[zG (v̇ +ur-wp)] (22) 𝑀𝑢̇ 0 𝑀𝑤̇ 0 𝑀𝑞 0
[0 𝑁𝑣̇ 0 𝑁𝑝̇ 0 𝑁𝑟̇ ]
 Pitching moment equation:
Because the symmetry at coordinates xz (right/ leftside thruster
∑ Mext =Iyy q̇ +(Ixx -Izz)pr+m[zG (u̇ +wq-vr)]-xG (ẇ +vp-uq) (23) symmetry). Substituting (29) into the notation of the Society of
Naval Architects and Marine Engineers (SNAME) to obtain
 Yawing moment equation: equation of forces and moments on the added mass:

∑ Next =Izz ṙ +(Iyy -Ixx )pq+m[zG (v̇ +ur-wp)] (24)  Surge added mass force:

C. Added Mass 𝑋𝐴 = 𝑋𝑢̇ 𝑢̇ + 𝑋𝑤̇ (𝑤̇ + 𝑢𝑞) + 𝑋𝑞̇ 𝑞̇ + 𝑍𝑤̇ 𝑤𝑞 + 𝑍𝑞̇ 𝑞2 − 𝑌𝑣̇ 𝑣𝑟

Added mass is understood as the pressure-induced forces and

−𝑌𝑝̇ 𝑟𝑝 − 𝑌𝑟̇ 𝑟 2 (30)
moments due to a forced motion of the body that is proportional
to the acceleration of the body. For a completely submerged
vehicle, the added mass coefficients are assumed constant and  Sway added mass force:
thus independent of the wave circular frequency [3]. Further
treatment on the subject of added mass can be found in 𝑌𝐴 = 𝑌𝑣̇ 𝑣̇ + 𝑌𝑝̇ + 𝑌𝑟̇ 𝑟̇ − 𝑋𝑤̇ (𝑢𝑝 − 𝑤𝑟) + 𝑋𝑢̇ 𝑢𝑟 − 𝑍𝑤̇ 𝑤𝑝
Newman [6]. The added mass force and moment equations can
be expressed in the form of +𝑍𝑞̇ 𝑝𝑞 + 𝑋𝑞̇ 𝑞𝑟 (31)

25
 Marine and Underwater Science and Technology 26

 Heave added mass force: To be precise damping coefficient hydrodynamic forces and
moments for the three coordinates of the objects is as follows
𝑍𝐴 = 𝑋𝑤̇ (𝑢̇ − 𝑤𝑞)𝑍𝑤̇ 𝑤̇ + 𝑍𝑞̇ 𝑞̇ − 𝑋𝑢̇ 𝑢𝑞 − 𝑋𝑞̇ 𝑞2 + 𝑌𝑣̇ 𝑣𝑝 𝑋𝐷 , 𝑌𝐷 , 𝑍𝐷 , 𝐾𝐷 , 𝑀𝐷 , 𝑁𝐷 where:

+𝑌𝑟̇ 𝑟𝑝 + 𝑌𝑝̇ 𝑝2 (32) 𝑋𝐷 = 𝑋𝑢|𝑢| 𝑢|𝑢| (37)

𝑌𝐷 = 𝑌𝑣|𝑣| 𝑣|𝑣| (38)

 Roll added mass moment:
𝑍𝐷 = 𝑍𝑤|𝑤| 𝑤|𝑤| (39)
𝐾𝐴 = 𝑌𝑝̇ 𝑣̇ + 𝐾𝑝̇ 𝑝̇ + 𝐾𝑟̇ 𝑟̇ + 𝑋𝑤̇ 𝑢𝑣 − (𝑌𝑣̇ − 𝑍𝑤̇ )𝑣𝑤

𝐾𝐷 = 𝐾𝑝|𝑝| 𝑝|𝑝| (40)

−(𝑌𝑟̇ + 𝑍𝑞̇ )𝑤𝑟 − 𝑌𝑝̇ 𝑤𝑝 − 𝑋𝑞̇ 𝑢𝑟 + (𝑌𝑟̇ + 𝑍𝑞̇ )𝑣𝑞

𝑀𝐷 = 𝑀𝑞|𝑞| 𝑞|𝑞| + 𝑀𝑤|𝑤| 𝑤|𝑤| (41)

+𝐾𝑟̇ 𝑝𝑞 − (𝑀𝑞̇ − 𝑁𝑟̇ )𝑞𝑟 (33)
𝑁𝐷 = 𝑁𝑟|𝑟| 𝑟|𝑟| + 𝑁𝑣|𝑣| 𝑣|𝑣| (42)
 Pitch added mass moment:

𝑀𝐴 = 𝑋𝑞̇ (𝑢̇ + 𝑤𝑞) + 𝑍𝑞̇ (𝑤̇ − 𝑢𝑞) + 𝑀𝑞̇ 𝑞̇ − 𝑋𝑤̇ (𝑢2 − 𝑤 2 )

E. Hydrostatic Forces and Moments
−(𝑍𝑤̇ − 𝑋𝑢̇ )𝑢𝑤 + 𝑌𝑝̇ 𝑣𝑟 − 𝑌𝑟̇ 𝑣𝑝 − 𝐾𝑟̇ (𝑝2 − 𝑟 2 ) The gravitational and buoyant forces are generally called
restoring forces in hydrodynamic terminology. The weight and
+(𝐾𝑝̇ − 𝑁𝑟̇ )𝑟𝑝 (34) buoyancy force vectors do not change with vehicle attitude for
bodies that are submerged. The expression of the Buoyant and
 Yaw added mass moment: gravitational force in earth fixed frame can be as: 𝐹𝑊 = 0𝐼 +
0𝐽 + 𝑊𝐾 and 𝐹𝐵 = 0𝐼 + 0𝐽 − 𝐵𝐾 . As mentioned above, the
buoyant and weight components are acting in the global vertical
𝑁𝐴 = 𝑌𝑟̇ 𝑣̇ + 𝐾𝑟̇ 𝑝̇ + 𝑁𝑟̇ 𝑟̇ − (𝑋𝑢̇ − 𝑌𝑣̇ )𝑢𝑣 − 𝑋𝑤̇ 𝑢𝑤
direction and they must be transformed into components in the
vehicle fixed frame in order to be added into the equations of
+(𝑋𝑞̇ + 𝑌𝑝̇ )𝑢𝑝 + 𝑌𝑟̇ 𝑢𝑟 + 𝑍𝑞̇ 𝑤𝑝 − (𝑋𝑞̇ + 𝑌𝑝̇ )𝑣𝑞
motion, the net vertical force components can be obtained as:
−(𝐾𝑝̇ − 𝑀𝑞̇ )𝑝𝑞 − 𝐾𝑟̇ 𝑞𝑟 (35) − sin 𝜃
𝐹ℎ𝑦𝑑𝑟𝑜𝑠𝑡𝑎𝑡𝑖𝑐 = (𝑊 − 𝐵) [ cos 𝜃 sin 𝜙 ] (43)
cos 𝜃 cos 𝜙
Added mass coefficient may be estimated using various
techniques such as the slender body theory [6] or strip theory
The weight part of the vertical force acts at the center of
[7]. The actual geometry of the ROV that is very complex such
as cabling systems, propellers. For modeling purposes, the gravity location of the vehicle, while the buoyancy part of the
ROV is assumed to have simple shapes such as tubes, balls, and vertical force acts at the center of buoyancy.
plate. Strip theory is used to calculate the coefficient of added
The resulting moment about the body center given by
mass since this technique is suitable for simple shapes assumed
by ROV. − sin 𝜃 − sin 𝜃
𝑀ℎ𝑦𝑑𝑟𝑜𝑠𝑡𝑎𝑡𝑖𝑐 = 𝑊𝜌𝐺 × [ cos 𝜃 sin 𝜙 ] − 𝐵𝜌𝐵 × [ cos 𝜃 sin 𝜙 ] (44)
D. Hydrodynamic Forces and Moments cos 𝜃 cos 𝜙 cos 𝜃 cos 𝜙
Hydrodynamics forces and moments associated with the
speed of the body and control surface deflections will be given This moment will be non-zero even if W and B are equal or
in this section. Damping of the ROV moves in 6 DOF at high 𝜌𝐺 and 𝜌𝐵 are zero. The center of gravity is needed to be
speed are non-linear and coupled. Assumptions of the ROV are located below the center of buoyancy considering the static
non-coupled motion. Reynolds number for ROV with a length stability of the vehicle. The total hydrostatic forces and
of 66 cm and a speed of 0.3 m/s is 4.25 x 105. The strength of moment’s vector can be written as:
forces and moments are equated to the right side of equation
(19-21) and (22-24) to get the full dynamics equations of 𝐹ℎ𝑦𝑑𝑟𝑜𝑠𝑡𝑎𝑡𝑖𝑐
[ ]=
motion used in the simulation [5]. 𝑀ℎ𝑦𝑑𝑟𝑜𝑠𝑡𝑎𝑡𝑖𝑐
−(𝑊 − 𝐵) sin 𝜃
The hydrodynamic damping force can be modeled as: (𝑊 − 𝐵) cos 𝜃 sin 𝜙
1 (𝑊 − 𝐵) cos 𝜃 cos 𝜙
𝐷(𝑢) = − 𝜌𝐶𝐷 𝐴|𝑢|𝑢 (36) (45)
2 (𝑦𝐺 𝑊 − 𝑦𝐵 𝐵) cos 𝜃 cos 𝜙 − (𝑧𝐺 𝑊 − 𝑧𝐵 𝐵) cos 𝜃 sin 𝜙
−(𝑥𝐺 𝑊 − 𝑥𝐵 𝐵) cos 𝜃 cos 𝜙 − (𝑧𝐺 𝑊 − 𝑧𝐵 𝐵) sin 𝜃
where CD, 𝜌, and A are drag coefficient, density (kg/m3), and [ (𝑥𝐺 𝑊 − 𝑥𝐵 𝐵) cos 𝜃 sin 𝜙 + (𝑦𝐺 𝑊 − 𝑦𝐵 𝐵) sin 𝜃 ]
frontal area (m2), respectively.
 Marine and Underwater Sci. and Tech., 2015, Vol. 1, No. 1

F. Forces and Moments of Thrusters  Roll (∑ 𝐾𝑒𝑥𝑡 ):

Frst, Flst, Ffwdt, Faftt are the thruster forces. The thruster forces 𝐼𝑥𝑥 𝑝̇ +(𝐼𝑧𝑧 − 𝐼𝑦𝑦 )𝑞𝑟
can also be obtained from the company of thruster or with
experiments [3]. Moment of thruster is multiplication of force −𝑚[𝑧𝐺 (𝑣̇ + 𝑢𝑟 − 𝑤𝑝)] = 𝑌𝑝̇ 𝑣̇ + 𝐾𝑝̇ 𝑝̇ + 𝐾𝑟̇ 𝑟̇ + 𝑋𝑤̇ 𝑢𝑣
by the distance from the center point to the thruster ROV. There
−(𝑌𝑣̇ − 𝑍𝑤̇ )𝑣𝑤 − (𝑌𝑟̇ + 𝑍𝑞̇ )𝑤𝑟
also needed to calculate the torque reaction of the thruster. To
be exact propulsion forces and moments on the ROV three −𝑌𝑝̇ 𝑤𝑝 − 𝑋𝑞̇ 𝑢𝑟 + (𝑌𝑟̇ + 𝑍𝑞̇ )𝑣𝑞
coordinate axes, namely, Xp, Yp, Zp, Kp, Mp, Np, where :
+𝐾𝑟̇ 𝑝𝑞 − (𝑀𝑞̇ − 𝑁𝑟̇ )𝑞𝑟+𝐾𝑝|𝑝| 𝑝|𝑝|
𝑋𝑝 = 𝐹𝑟𝑠𝑡 + 𝐹𝑙𝑠𝑡 (46)
−(𝑧𝐺 𝑊 − 𝑧𝐵 𝐵) cos 𝜃 sin 𝜙 (56)
𝑌𝑝 = 0 (47)
 Pitch (∑ 𝑀𝑒𝑥𝑡 ):
𝑍𝑝 = 𝐹𝑓𝑤𝑑𝑡 + 𝐹𝑎𝑓𝑡𝑡 (48)
𝐼𝑦𝑦 𝑞̇ + (𝐼𝑥𝑥 − 𝐼𝑧𝑧 )𝑝𝑟 + 𝑚[𝑧𝐺 (𝑢̇ + 𝑤𝑞– 𝑣𝑟)
𝐾𝑝 = 0 (49)
−𝑥𝐺 (𝑤̇ + 𝑣𝑝 – 𝑢𝑞)] = 𝑋𝑞̇ (𝑢̇ + 𝑤𝑞) + 𝑍𝑞̇ (𝑤̇ − 𝑢𝑞)
𝑀𝑝 = (𝐹𝑟𝑠𝑡 + 𝐹𝑙𝑠𝑡 )𝑧𝑝 + 𝐹𝑎𝑓𝑡𝑡 𝑥𝑝𝑎𝑓𝑡 − 𝐹𝑓𝑤𝑑𝑡 𝑥𝑝𝑓𝑤𝑑 (50)
+𝑀𝑞̇ 𝑞̇ − 𝑋𝑤̇ (𝑢2 − 𝑤 2 ) − (𝑍𝑤̇ − 𝑋𝑢̇ )𝑢𝑤
𝑁𝑝 = (𝐹𝑟𝑠𝑡 − 𝐹𝑙𝑠𝑡 )𝑦𝑝 (51)
+𝑌𝑝̇ 𝑣𝑟 − 𝑌𝑟̇ 𝑣𝑝 − 𝐾𝑟̇ (𝑝2 − 𝑟 2 )
G. Mathematical Modelling for Simulation of ROV
+(𝐾𝑝̇ − 𝑁𝑟̇ )𝑟𝑝 + 𝑀𝑞|𝑞| 𝑞|𝑞|
The simulation of ROV was made possible by using the
above set of equations and employing pertinent parameters. +𝑀𝑤|𝑤| 𝑤|𝑤| − (𝑥𝐺 𝑊 − 𝑥𝐵 𝐵) cos 𝜃 cos 𝜙
External force was denoted as ∑ 𝑋𝑒𝑥𝑡 representing the sum of
force associated with additional mass, hydrodynamic damping −(𝑧𝐺 𝑊 − 𝑧𝐵 𝐵) sin 𝜃 + (𝐹𝑟𝑠𝑡 + 𝐹𝑙𝑠𝑡 )𝑧𝑝
force, restoring force and propulsion force (52). External forces +𝐹𝑎𝑓𝑡𝑡 𝑥𝑝𝑎𝑓𝑡 − 𝐹𝑓𝑤𝑑𝑡 𝑥𝑝𝑓𝑤𝑑 (57)
for surge, sway and heave directions, and external moments on
roll, pitch and yaw directions can be formulated following the  Yaw (∑ 𝑁𝑒𝑥𝑡 ):
same principle. Overall equations can be summarized as
𝐼𝑧𝑧 𝑟̇ + (𝐼𝑦𝑦 − 𝐼𝑥𝑥 )𝑝𝑞
follows:
∑ 𝑋𝑒𝑥𝑡 =𝑋𝐴 + 𝑋𝐷 + 𝑋𝑅 + 𝑋𝑃 (52) +𝑚[𝑥𝐺 (𝑣̇ + 𝑟𝑢– 𝑝𝑤)] = 𝑌𝑟̇ 𝑣̇ + 𝐾𝑟̇ 𝑝̇ + 𝑁𝑟̇ 𝑟̇

 Surge (∑ 𝑋𝑒𝑥𝑡 ): −(𝑋𝑢̇ − 𝑌𝑣̇ )𝑢𝑣 − 𝑋𝑤̇ 𝑢𝑤 + (𝑋𝑞̇ + 𝑌𝑝̇ )𝑢𝑝

𝑚{𝑢̇ + 𝑤𝑞 − 𝑣𝑟 − 𝑥𝐺 (𝑞 2 + 𝑟 2 ) +𝑌𝑟̇ 𝑢𝑟 + 𝑍𝑞̇ 𝑤𝑝 − (𝑋𝑞̇ + 𝑌𝑝̇ )𝑣𝑞

+ 𝑧𝐺 (𝑝𝑟 + 𝑞̇ )} = 𝑋𝑢̇ 𝑢̇ + 𝑋𝑤̇ (𝑤̇ + 𝑢𝑞) + 𝑋𝑞̇ 𝑞̇ + 𝑍𝑤̇ 𝑤𝑞 −(𝐾𝑝̇ − 𝑀𝑞̇ )𝑝𝑞 − 𝐾𝑟̇ 𝑞𝑟 + 𝑁𝑟|𝑟| 𝑟|𝑟|
+𝑍𝑞̇ 𝑞 2 − 𝑌𝑣̇ 𝑣𝑟 − 𝑌𝑝̇ 𝑟𝑝 − 𝑌𝑟̇ 𝑟 2 +𝑁𝑣|𝑣| 𝑣|𝑣| + (𝑥𝐺 𝑊 − 𝑥𝐵 𝐵) 𝑐𝑜𝑠 𝜃 𝑠𝑖𝑛 𝜙
+𝑋𝑢|𝑢| 𝑢|𝑢| − (𝑊 − 𝐵) sin 𝜃 + 𝐹𝑟𝑠𝑡 + 𝐹𝑙𝑠𝑡 (53) +(𝐹𝑟𝑠𝑡 − 𝐹𝑙𝑠𝑡 ) (58)
 Sway (∑ 𝑌𝑒𝑥𝑡 ): III. PID CONTROLLER AND SIMULATION

𝑚{𝑣̇ − 𝑤𝑝 + 𝑢𝑟 + 𝑥𝐺 (𝑝𝑞 + 𝑟̇ ) A. PID Controller for Pitch Movement

In the simulation, there is no control system that maintains
+ 𝑧𝐺 (𝑞𝑟 − 𝑝̇ )} = 𝑌𝑣̇ 𝑣̇ + 𝑌𝑝̇ + 𝑌𝑟̇ 𝑟̇ − 𝑋𝑤̇ (𝑢𝑝 − 𝑤𝑟)
the hovering state, in surge and heave motion. The influence of
+ 𝑋𝑢̇ 𝑢𝑟 pitch angle greatly affects the simulation results. Therefore, in
−𝑍𝑤̇ 𝑤𝑝 + 𝑍𝑞̇ 𝑝𝑞 + 𝑋𝑞̇ 𝑞𝑟 + 𝑌𝑣|𝑣| 𝑣|𝑣| order to maintain the hovering state, the pitch angle should be
controlled. The value of PID control in MATLAB/Simulink is
+(𝑊 − 𝐵) cos 𝜃 sin 𝜙 (54) determined by adding PID controller blocks in the model in
 Heave (∑ 𝑍𝑒𝑥𝑡 ): order to make a close loop system [8][9][10]. Figure 3 shows
the addition of PID block diagram into ROV control loop.
𝑚{𝑤̇ − 𝑢𝑞 + 𝑣𝑝 + 𝑥𝐺 (𝑟𝑝 − 𝑞̇ )
Inside the PID controller block, a Function Block
− 𝑧𝐺 (𝑞 2 + 𝑝2 )} = 𝑋𝑤̇ (𝑢̇ − 𝑤𝑞) + 𝑍𝑤̇ 𝑤̇ + 𝑍𝑞̇ 𝑞̇ − 𝑋𝑢̇ 𝑢𝑞 Parameters PID Controller as shown in Figure 4 is embedded.
The PID linearized value is tuned automatically by the toolbox.
−𝑋𝑞̇ 𝑞 2 + 𝑌𝑣̇ 𝑣𝑝 + 𝑌𝑟̇ 𝑟𝑝 + 𝑌𝑝̇ 𝑝2 The obtained PID controller’s parameters are:
+𝑍𝑤|𝑤| 𝑤|𝑤| + (𝑊 − 𝐵) cos 𝜃 cos 𝜙  Proportional -60.1528757451922
+𝐹𝑓𝑤𝑑𝑡 + 𝐹𝑎𝑓𝑡𝑡 (55)  Integral -12.3345980314996
 Derivative -26.6314325750794.

27
 Marine and Underwater Science and Technology 28

Figure 3 Additions to the PID block diagram hovering condition,

surge, and heave motion

Figure 5 PID tuner

Figure 6 Modeling of the ROV in MATLAB Simulink and Virtual

Reality toolbox
Figure 4 Function block parameter of PID controller in hovering
condition for surge and heave motion
IV. RESULT AND DISCUSSION
Figure 5 shows PID tuner quiescent conditions, which The current study was limited to the longitudinal motion of
automatically displays the value in response time of 0.0767 s, the ROV showing the time history of pitch angle, forward and
transient behavior damping value of 0.6, the result of PID vertical velocities for hovering, surge, and heave motions. To
parameter values are rise time of 0.0481 s, settling time of 5.02 analyze the stability of the ROV, the equations of motion were
s, overshoot of 13.2 %, peak of 1.13, gain margin of -24.7 dB, linearized around pre-defined trim conditions. Eigen value
phase margin of 60 deg, and closed-loop stability as stable. analysis was then performed for the system matrix obtained
from the linearized model. TABLE II shows the eigenvalue
B. Dynamic System Modeling of ROV Using MATLAB structure for the open loop system. The existence of positive
Simulink real part of eigenvalues indicates that the system without
In this research, dynamic system of ROV is modeled using feedback is unstable. Figure 7 shows the corresponding
MATLAB 2013b/Simulink that is connected with toolbox sl3d root-locus diagram of the open-loop system.
program, and the 3D representation is shown with virtual
reality toolbox. All dynamic model of the ROV is built with A. Dynamic Simulation of ROV in Hovering Condition
interconnection block that represent mathematic analysis of The objective of hover control is to eliminate or minimize the
ROV. Figure 6 shows the dynamic model of the ROV. ROV angular and linear velocities thus enable the ROV to maintain
dynamics shows physical condition of ROV that gives the state its stationary position. Figure 8 indicates that the pitch angle is
of position, velocity and acceleration, both linear and angular. negligibly small throughout the period of analysis, the
 Marine and Underwater Sci. and Tech., 2015, Vol. 1, No. 1

maximum pitch angle being 16 x 10-3 deg. Further, Figure 9

-3
x 10
16

shows that the forward displacement and vertical velocity were 14

also very small. The ROV only travelled 0.033 m from its 12
initial point within 50 secs.
10

The maximum vertical velocity is 3.3 × 10-4 m/s. Overall 8

 (deg)
results indicate that the feedback control is effective in 6

maintaining the ROV in the hovering condition. 4

B. Dynamic Simulation of ROV in Heave Motion 2

0
The controlled heaving motion of the ROV is characterized
-2
by vertical displacement with negative pitch angle. Figure 10 0 5 10 15 20 25 30 35 40 45 50
Time (s)
shows that the ROV starts its vertical motion pitching down. Figure 8 Orientation of the pitch angle in hovering condition
This is an indication that the system is slightly non-minimum
x Position vs Time
phase. The maximum pitch angle was -1.8 degree and it 0.035

achieved a constant pitch angle of 0 degree within 25 s. 0.03

Figure 11 shows that the ROV travels 12.5 m vertically 0.025

within the above time frame with the maximum speed of 0.25
0.02
m/s which is achieved within 5s from its initial zero velocity.

x (m)
The steady state velocity is achieved quickly without an 0.015

overshoot. The control performance for the heave motion is 0.01

overall satisfactory.
0.005

TABLE II EIGENVALUE, DAMPING, AND FREQUENCY IN OPEN LOOP 0

0 5 10 15 20 25 30 35 40 45 50
Time (s)
Eigenvalue Damping Freq. (rad/s) -4
w Velocity vs Time
x 10
0
0.00e+00 -1.00E+00 0.00e+00
0.00e+00 -1.00E+00 0.00e+00
0.00e+00 -1.00E+00 0.00e+00
-1
0.00e+00 -1.00E+00 0.00e+00
-1.80e+00 1.00E+00 1.80e+00
w (m/s)

1.80e+00 -1.00E+00 1.80e+00 -2

-5.68e-06 1.00E+00 5.68e-06

-1.05e-04 1.00E+00 1.05e-04
-3
1.48e+00 -1.00E+00 1.48e+00
-1.48e+00 1.00E+00 1.48e+00
0 5 10 15 20 25 30 35 40 45 50
Time (s)
-4.35e-05 1.00E+00 4.35e-05
-1.34e-04 1.00E+00 1.34e-04
Figure 9 Position and velocity in hovering condition
0.2
Pole-Zero Map
1
0.86 0.76 0.58 0.35 0
0.92
0.8
-0.2
0.6
0.96
-0.4
0.4
0.984
System: sys System: sys
-0.6
Imaginary Axis (seconds-1)

Pole : -0.000134 Pole : 1.8

 (deg)

0.2
0.996 Damping: 1 Damping: -1
Overshoot (%): 0
Frequency (rad/s): 0.000134
Overshoot (%): 0
Frequency (rad/s): 1.8
-0.8
2 1.75 1.5 1.25 1 0.75 0.5 0.25
0
System: sys
Pole : 1.48 -1
0.996 Damping: -1
-0.2 Overshoot (%): 0
Frequency (rad/s): 1.48

0.984
-1.2
-0.4

0.96
-1.4
-0.6

-1.6
-0.8
0.92

0.86 0.76 0.58 0.35 -1.8

-1
-2 -1.5 -1 -0.5 0 0.5 1 1.5 2
0 5 10 15 20 25 30 35 40 45 50
Real Axis (seconds -1) Time (s)

Figure 10 Orientation of the pitch angle in heave motion

Figure 7 Root location of the open loop system

29
 Marine and Underwater Science and Technology 30

C. Dynamic Simulation of ROV in Surge Motion 50

x Position vs Time

The surge motion of the ROV was typified by an increased 45

pitch angle within the first 4s followed by monotonically 40
decreasing angle reaching zero after 25 s. The maximum pitch
35
angle was 12 degree as depicted in Figure 12.
30
Figure 13 shows that the ROV travels 47 m during 50s with

x (m)
25
step forward velocity profile. The steady state forward velocity
20
of 0.92 m/s was achieved within 15s with a slight overshoot. As
15
in hover and heave modes, the control for the surge motion is
effective. We can conclude that the ROV is fully controllable in 10

its longitudinal plane. Further study needs to be performed for 5

evaluating its controllability in the lateral directional mode. We 0

0 5 10 15 20 25 30 35 40 45 50
envisage the study of other control techniques application for Time (s)
the ROV including the Coefficient Diagram Method (CDM), u Velocity vs Time
1
studied earlier in [11][12], Model Predictive Control as in [13],
Linear Parameter Varying (LPV) [14] and other techniques 0.9

surveyed in [15]. The control performance will be analyzed and 0.8

evaluated within the Hardware In the Loop (HIL) simulation 0.7

environment [16]. 0.6

u (m/s)
z Position vs Time
14 0.5

0.4
12
0.3
10
0.2

8 0.1
z (m)

0
6 0 5 10 15 20 25 30 35 40 45 50
Time (s)

4
Figure 13 Position and velocity in surge motion
2

0
V. CONCLUSIONS
0 5 10 15 20 25 30 35 40 45 50
Time (s) The work is focused on the modeling, simulation and control
w Velocity vs Time
0.25
of the ROV. The ROV was first designed by using the
SolidWorks optimizing the balance between the gravitational
0.2
and buoyancy forces to enable the vehicle to float. The
0.15
mathematical model of the ROV was developed based on
w (m/s)

Euler-Newton equations leading to a set of nonlinear

0.1 differential equations. The nonlinear simulation was realized
using MATLAB 2013b/Simulink including its three
0.05
dimensional virtual reality environment. Stability was analyzed
0
from linearized models around pre-defined trim trajectories
0 5 10 15 20 25 30 35 40 45 50
Time (s) which include hover, heave and surge motion. PID control
Figure 11 Position and velocity in heave motion design was performed based on the models and its performance
12
was evaluated. Overall analysis demonstrates that the feedback
10
control was effective for manipulating the ROV in the surge
and heave motions as well as maintaining hovering position.
8
Further study is envisaged for evaluating various control
techniques including its implementation in the Hardware In the
 (deg)

6
Loop (HIL) simulation environment.
4

ACKNOWLEDGEMENTS
2

The authors would like to be obliged to Diponegoro

0
0 5 10 15 20 25 30 35 40 45 50 University for providing laboratory facilities.
Time (s)

Figure 12 Orientation of pitch angle in surge motion

 Marine and Underwater Sci. and Tech., 2015, Vol. 1, No. 1

REFERENCES
[1] Alim, A.S., 2008, Evaluasi Kebijakan Dalam Rangka Implementasi
Konvensi Hukum Laut International (UNCLOS 1982) di Indonesia,
Departemen Kelautan dan Perikanan Sekertariat Jenderal Satuan Kerja
Dewan Kelautan Indonesia, Jakarta, Indonesia.
[2] Chandra,Y., Purnomo, D.K., Suryawati N, E., 2011, Rancang Bangun
Sistem Ballast Pada ROV (Remotely Underwater Vehicle), Jurusan
Teknik Mekatronika-Politeknik Elektronika Negeri Surabaya, Institut
Teknologi Sepuluh Nopember (ITS) Surabaya.
[3] Fossen, T.I., 1994, Guidance and Control of Ocean Vehicles, John Wiley
and Sons.
[4] Geridonmez, F., 2007, Simulation Of Motion Of An Underwater Vehicle,
Master Thesis, Aerospace Engineering Department, Middle East
Technical University.
[5] Chuan, T.S., 1999, Modeling and Simulation of the Autonomous
Underwater Vehicle, Autolycus, Master Thesis, Department of Ocean
Engineering, Massachusetts Institute of Technology, USA.
[6] J. N. Newman. 1997. Marine Hydrodynamics. Massachusetts Institute of
Technology Press.
[7] S. H. Hoerner. 1965. Fluid Dynamic Lift. Hoerner Fluid Dynamics.
[8] Lin F. 2007. Robust Control Design: An Optimal Control Approach. John
Wiley Sons. England. CrossRef
[9] Budiyono, A. et al. Intelligent Unmanned Systems: Theory and
Applications. Springer-Verlag Berlin Heidelberg. 2009. CrossRef
[10] Siong Chin, Cheng. 2013. Computer-Aided Control Systems Design.
New York: CRC Press Taylor and Francis Group.
[11] Agus Budiyono, Muljowidodo Kartidjo and Agus Sugama, Coefficient
Diagram Method for the Control of an Unmanned Underwater Vehicle, ,
Indian J Mar Sci., 38(3):316-323, Sept. 2009
[12] Budiyono, A. "Robust Control Synthesis for an Unmanned Underwater
Vehicle." World Academy of Science, Engineering and Technology 60
(2009): 48-55
[13] Budiyono, Agus. "Model predictive control for autonomous underwater
vehicle." Indian Journal of Marine Sciences 40.2 (2011): 191
[14] H.Y. Sutarto and Agus Budiyono, Development of linear parameter
varying control system for autonomous underwater vehicle, Indian J
Geo-Mar Sci., 40(2):275-286, Apr. 2011
[15] Budiyono, A., Advances in Unmanned Underwater Vehicles
Technologies, Indian J Mar Sci., 38(3):282-295, Sept. 2009
[16] Kaliappan, VK, Budiyono, A and Min, D. "Hardware-In-the-loop
simulation platform for the design, testing and validation of autonomous
control system for unmanned underwater vehicle." Indian Journal of
Geo-Marine Sciences 41.6 (2012): 575-580.

31